whatmodel.sh / T0 3B
T0 3B
↗ share
+ contribute
T0 3B is an open-weights language model from BigScience, released October 2021. Its published size is 3B parameters. Licence: apache-2.0.
Catalogue identity notes
BigScience T0_3B (Sanh et al., 'Multitask Prompted Training Enables Zero-Shot Task Generalization', arXiv 2110.08207, 2021-10-15); public checkpoint bigscience/T0_3B (Apache-2.0).
Model information sources
Score within the LMArena (text) comparison set only.
$ how it compares
The LLMs ranked either side of T0 3B, out of 877 with a score, then more from BigScience.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 823 |  StableLM-Base-Alpha 3B v2 | 24±1 | local | ||
| 824 |  OPT-6.7B | 24±1 | local | ||
| 825 | T0 3B | 23±4 | comparison only | ||
| 826 | BLOOMZ 1.7B | 23±3 | comparison only | ||
| 827 |  Qwen1.5 0.5B Chat | 23±3 | local | ||
| more from BigScience | |||||
| 743 | T0 11B | 35±3 | comparison only | ||
| 759 | BLOOMZ 176B | 33±2 | comparison only | ||
| 796 | BLOOMZ 7.1B | 29±2 | comparison only | ||
$ what it's good at
other tests3 not in a task · 93% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| WebQSP closed-book zero-shot (answer-in-generation accuracy) | 21.4% ✎ | ██···██····█ beats 1/4 | 50% |
| BIG-Bench Hard, 27 subsets, zero-shot direct (CoT Collection protocol) | 27.5% ✎ | █····█··█··█ beats 0/3 | 26% |
| Mintaka (English) closed-book zero-shot (answer-in-generation accuracy) | 14.8% ✎ | ▄····▄···▄·█ beats 0/4 | 17% |
$ where to run it
No verified API or download recorded yet · add one
Official weights and documentation: huggingface.co/bigscience ↗
$ how the score was worked out · Knowledge-Augmented Language… 67%, The CoT Collection: Improving… 26%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Knowledge-Augmented Language Model Prompting for Zero-Shot Knowledge Graph Question Answering — authors' own runs | 2 | 67% |
| The CoT Collection: Improving Zero-shot and Few-shot Learning of Language Models via Chain-of-Thought Fine-Tuning — authors' own runs | 1 | 26% |
| prior · pull toward the middle | 7% | |
| weighted average | = 23.3 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json